Cassandra 1.0 actually has a feature that, if the documentation is to be believed, relieves it of its worst misfeature: casually dropped replication events. The story of the fix is entertainingly told in https://issues.apache.org/jira/browse/CASSANDRA-2034 (if you're into a certain kind of entertainment).
Thanks to Jake Luciani for the clue. I don't understand why none of the Cassandra apologists who tried to tell me I was all wet about its replication flaws didn't actually, you know, mention this before now. But it's not the biggest mystery I've ever seen.
Anyway, without having tried it 1.0, I can't give it a thumbs up. Still, kudos for adding a feature that looks like the right thing. I hope it really is.
PS: Cassandra needs to stop requiring half your disk to stay empty, but at least they're up front about that requirement. If you can afford lots of unused disk, knock yourself out.
FYI, we also addressed the 50%-of-disk-space worst case scenario in 1.0: http://www.datastax.com/dev/blog/leveled-compaction-in-apache-cassandra. With leveled compaction, the free space requirement is 10x the (fixed) sstable size, which defaults to 5MB.
Posted by: Jonathan Ellis | December 14, 2011 at 12:17 PM